CN108984643B - Sport place extraction method based on jogging GPS track data - Google Patents
Sport place extraction method based on jogging GPS track data Download PDFInfo
- Publication number
- CN108984643B CN108984643B CN201810652492.0A CN201810652492A CN108984643B CN 108984643 B CN108984643 B CN 108984643B CN 201810652492 A CN201810652492 A CN 201810652492A CN 108984643 B CN108984643 B CN 108984643B
- Authority
- CN
- China
- Prior art keywords
- track
- jogging
- data
- gps
- trajectory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses a method for extracting a sport place based on jogging GPS track data, which comprises the following steps: step 1, acquiring original jogging GPS trajectory data from a database, decomposing the original jogging GPS trajectory data into two-dimensional time series curve data, and extracting a key point set from the two-dimensional time series curve data; step 2, constructing a triangular matrix at the sub-track distance according to the track point set corresponding to the key point set, and searching a winding sub-track meeting the winding track mode characteristics according to a line scanning search matrix; outputting a group winding sub-trajectory line set after traversing all GPS trajectory lines in the database; and 3, extracting geometric information and semantic information of the motion place from the group circle trajectory line set. The method can quickly and automatically extract the jogging circle movement place information on the semantic level, reduces the cost for acquiring the fine movement place information, and is simple and easy to realize.
Description
Technical Field
The invention relates to the technical field of geographic information and spatiotemporal trajectory data mining, in particular to a jogging GPS trajectory data-based sport place extraction method.
Background
Under the influence of the national body-building trend, the health consciousness of people is increasingly strengthened, more and more people pay attention to physical exercise in daily life, and the living idea of pursuing health through exercise is deep in mind. Jogging is an important mode of sports, fitness, leisure and entertainment, and has important effects of strengthening physique, losing weight, preventing obesity, relieving pressure and the like. The circle jogging is used as a main behavior in jogging, and the significance of accurately identifying a circle behavior mode and extracting circle motion place information is great. How to quickly and accurately acquire jogging place information (geometric information and semantic information) in a city and provide decisions and suggestions for jogging place recommendation, jogging path planning, runner recommendation and the like is a research topic with high practical value. The traditional place information extraction method comprises manual field investigation measurement and remote sensing technology. The manual survey measurement mode has high cost and long updating period, and is difficult to meet the requirements of practical application. The high-resolution remote sensing image technology is mainly used for extracting urban land utilization data, monitoring land utilization coverage change and the like, but the method cannot extract site information of specific human activities (such as jogging and riding) and cannot sense dynamic changes of human activity behaviors and activity site characteristic information.
At present, the wide application of software and hardware such as mobile positioning devices (GPS, bracelet), sports apps (dingdong, yue race circle) and the like generates massive leisure sports track data (jogging, riding and the like) and sports attribute data (energy consumption, heart rate and the like) with individual marks. The high-space-time-resolution behavior data is helpful for mining human activity behavior patterns, analyzing coupling relations between places and behaviors, sensing characteristics of the activity places and monitoring body health changes. The space-time trajectory has the advantages of low cost, high occurrence, high space-time resolution, big data, wide sources and the like, and is widely applied to extraction of relevant urban information. The method mainly comprises the following steps: the method comprises the following steps of road information extraction, city hotspot analysis, human behavior pattern detection, spatial data acquisition and updating and the like. However, it has been studied to detect land use types and identify urban functional areas by using social media data, mobile phone call data, taxi GPS track data, and the like. The researches mostly use the traffic district as the unit for land semantic recognition, and the research on the information extraction of the activity places from the view point of human behaviors is less. Through analysis of existing related researches, the fact that GPS track data are used for extracting the information of the human jogging circle motion place is still a blank research.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method for extracting a sport place based on jogging GPS track data aiming at the defects in the prior art.
The technical scheme adopted by the invention for solving the technical problems is as follows:
the invention provides a jogging GPS trajectory data-based sport place extraction method, which comprises the following steps:
step 1, acquiring original jogging GPS trajectory data from a database, decomposing the original jogging GPS trajectory data into two-dimensional time series curve data, and extracting a key point set from the two-dimensional time series curve data;
step 2, constructing a triangular matrix at the sub-track distance according to the track point set corresponding to the key point set, and searching a winding sub-track meeting the winding track mode characteristics according to a line scanning search matrix; outputting a group winding sub-trajectory line set after traversing all GPS trajectory lines in the database;
and 3, extracting geometric information and semantic information of the motion place from the group circle trajectory line set.
Further, the specific method of step 1 of the present invention is:
step 1.1, determining a parameter threshold, comprising: shortest path distance, maximum direction distance, minimum cycle time;
step 1.2, inputting original jogging GPS trajectory data, and decomposing original space-time trajectory coordinates (x, y, t) into two-dimensional time sequence curve data of (t, x) and (t, y);
step 1.3, taking a starting point, an end point, a maximum value and a minimum value point in the two-dimensional time sequence data as key points; merging the key point sets extracted from the (t, x) and (t, y) sequences to obtain a whole key point set KP ═ k1,k2,...,kn}; extracting original track points corresponding to the key points according to the time sequence to be used as a search point set SP ═ s of the winding behavior sub-track1,s2,...,sn}。
Further, the parameter threshold determined in step 1.1 of the present invention is specifically:
the shortest path distance minPath is 400m, the maximum direction distance maxDirct is 16m, and the minimum cycle time minTime is 5 min.
Further, the method for selecting the extreme point in step 1.3 of the present invention specifically comprises:
for a sequence of three points (x)i-1,xi,xi+1) If point xiSatisfy { xi<xi-1&&xi<xi+1Either { x } ori>xi-1&&xi>xi+1The pixel is a local extreme point; extreme point xiThe time period of the extreme value is kept, namely the time period of the point and the front and rear extreme points is greater than minTime/4, and the minTime represents the minimum cycle time; otherwise, the point is taken as a redundant key point and deleted.
Further, the specific method of step 2 of the present invention is:
step 2.1, using the SP point set obtained in the step 1.3 as a row according to the time sequenceThe method comprises the steps that a triangular matrix on a track distance is constructed in a row mode, and each matrix unit records the path distance (pathDist), the direction distance (dirDist) and the winding duration (T) of a sub-track formed by track points corresponding to the row and the columndur) (ii) a For any trajectory T ═ pn,pn+1,...,pmThe calculation formulas of the track path distance, the direction distance, the duration and the like are as follows:
dirDist(T)=Dist(pn,pm)
Tdur=T(pm)-T(pn)
wherein Dist (p)k,pk+1) Euclidean distance, T (p), representing two trace pointsk) Representing points of track pkThe time of recording;
step 2.2, searching the triangular matrix on the track distance according to rows, and if 3 eigenvalues in the matrix unit all meet the track characteristics of the winding behavior, the conditions are met:
{pathDist≥minPath&&dirDist≤maxDirct&&Tdur≥minTime}
the sub-track corresponding to the row and column is a winding sub-track, and the winding sub-tracks are sequentially searched according to rows by the matrix row where the terminal point of the sub-track is located; if the winding sub-track is searched, repeatedly executing the step 2.2 to perform depth-first search; outputting a winding action sub-track until the track line is processed completely without a winding action sub-track;
and 2.3, repeatedly executing the step 2.1 to the step 2.2 to traverse all the GPS track lines, and finally obtaining a group circle behavior track set.
Furthermore, in step 3 of the invention, geometric information and semantic information of the motion place are extracted from the group circle trajectory line set by using a Delaunay triangulation network and an inverse address coding method.
Further, the specific method of step 3 of the present invention is:
3.1, constructing a constrained Delaunay triangulation network for the group winding trajectory line set extracted in the step 2.3, and deleting the integral long edge in the Delaunay triangulation network, so that winding trajectory lines of each motion place are gathered into one type; the overall long-edge deletion threshold globalarue is calculated as follows:
GlobalValue=Mean(DT)+α×Variation(DT)
wherein mean (DT) represents the average value of all side lengths of the triangular net DT; variance (DT) represents the standard deviation of all side lengths of the triangular net; alpha is a regulating parameter and is 1 by default; for any triangle side, if the side length is larger than the GlobalValue, deleting the side;
step 3.2, after deleting the local long edge of the cluster obtained in the step 3.1 again, merging triangles for each cluster to extract polygons, and simplifying the polygons to obtain a motion path polygon; the local long edge deletion threshold LocalValue is calculated as follows:
wherein the content of the first and second substances,representing a cluster GiMidpoint pjThe average value of all side lengths in the second-order neighborhood of (1); LocalVarioration (p)j) Representing a cluster GiNeutralization point pjThe standard deviation of the side length of the direct connection side; beta is a regulating parameter, and the default is 2;
step 3.3, deleting an inner ring of the path polygon extracted in the step 3.2 to extract a circle place polygon, and then extracting circle motion place semantic information by utilizing Baidu reverse address coding, wherein the semantic types comprise: stadiums, living quarters, parks, roads, lake greens, leisure greens, greens squares.
Further, the database of step 1 of the present invention records: the mass jogging GPS trajectory line data with individual marks is generated by the mobile positioning equipment and the sport APP.
The invention has the following beneficial effects: the invention relates to a method for extracting a sport place based on jogging GPS track data, which constructs a method for acquiring sport place information from the jogging GPS track data by mining the jogging GPS track data; the method can quickly identify and extract the sub-track of the circle behavior from the massive jogging GPS track lines, and extract the jogging place information by applying the group circle track line set on the semantic level, thereby filling the blank of the research of the jogging place extraction method; meanwhile, the invention takes the crowd-sourced GPS track as a data source, reduces the extraction cost of the sport place information, and has simple detection and extraction method and easy realization.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a flow chart of an embodiment of the present invention;
fig. 2 is a schematic diagram of the extraction of a key point set of a time series curve according to an embodiment of the present invention, in which fig. 2(a) is a schematic diagram of the decomposition of an original GPS trajectory line into two-dimensional time series data, and fig. 2(b) is a schematic diagram of the extraction of a key point set of the series data;
FIG. 3 is a schematic diagram of the construction of the trajectory distance matrix and the search of the trajectory distance matrix for the winding behavior sub-trajectories according to the embodiment of the present invention;
FIG. 4 is a schematic diagram of a group jogging circle behavior trajectory extraction result according to an embodiment of the invention;
fig. 5 is a schematic diagram of extracting a motion place by constraining a Delaunay triangulation network according to an embodiment of the present invention, where fig. 5(a) is a schematic diagram of constructing a Delaunay triangulation network for a group circle trajectory set, and fig. 5(b) is a schematic diagram of deleting an overall long-edge clustering trajectory;
fig. 6 is a schematic diagram and a result of extracting motion location information according to an embodiment of the present invention, in which fig. 6(a) is a schematic diagram of extracting a circle motion location path polygon, and fig. 6(b) is a schematic diagram of extracting a jogging motion location polygon and location semantic information.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the method for extracting a sport area based on jogging GPS track data according to the embodiment of the present invention includes the following steps:
step 1: as shown in fig. 2, decomposing the latitude and longitude coordinates of the original jogging GPS trajectory into time series data of (t, x) and (t, y), and extracting a key point set for the time series curve;
step 1.1: determining a parameter threshold: the minimum path distance minPath is 400m, the maximum direction distance maxDerct is 16m, and the minimum cycle time minTime is 5 min;
step 1.2: as shown in fig. 2(a), inputting a jogging GPS track line, decomposing the original space-time track line coordinates (x, y, t) into two-dimensional time series data of (t, x), (t, y);
step 1.3: as shown in fig. 2(b), the start point, the end point, the maximum value, and the minimum value point in the time-series data are set as key points. For a sequence of three points (x)i-1,xi,xi+1) If point xiSatisfy { xi<xi-1&&xi<xi+1Either { x } ori>xi-1&&xi>xi+1And the pixel is a local extreme point. Extreme point xiThe time period for maintaining the extremum (i.e. the time period between the point and the preceding and following extremum points) should be greater than minTime/4, otherwise the point is deleted as the redundant key point. Merging the key point sets extracted from the (t, x) and (t, y) sequences to obtain a whole key point set KP ═ k1,k2,...,kn}. Extracting original track points corresponding to the key points according to the time sequence to be used as a search point set SP ═ s of the winding behavior sub-track1,s2,...,snResults are shown in FIG. 2 (b);
step 2: according to the key point set extracted in the step 1.3, constructing a triangular matrix on the sub-track distance for the track point set corresponding to the key point set, and searching a winding sub-track meeting the winding track mode characteristics according to a line scanning search matrix; outputting a group winding sub-trajectory line set after traversing all GPS trajectory lines in the database;
step 2.1: as shown in fig. 3Constructing a track distance upper triangular matrix by taking the SP point set obtained in the step 1.3 as a row and a column according to a time sequence, and recording the path distance (pathDist), the direction distance (dirDist) and the winding time length (T) of a sub-track formed by track points corresponding to the row and the column by each matrix unitdur). For any trajectory T ═ pn,pn+1,...,pmThe calculation formulas of the track path distance, the direction distance, the duration and the like are as follows:
dirDist(T)=Dist(pn,pm)
Tdur=T(pm)-T(pn)
dist (p) in the above formulak,pk+1) Euclidean distance, T (p), representing two trace pointsk) Representing points of track pkThe time of recording;
step 2.2: as shown in fig. 3, the trajectory distance matrix is searched in rows, and if all 3 eigenvalues in the matrix unit satisfy the circle behavior trajectory characteristic, the condition is satisfied:
{pathDist≥minPath&&dirDist≤maxDirct&&Tdur≥minTime}
the sub-track corresponding to the row and column is a winding sub-track, and the winding sub-track is sequentially searched according to the rows by the matrix row where the terminal point of the sub-track is located. If the winding sub-track is searched, depth-first searching is carried out according to the steps. Outputting a winding action sub-track until the track line is processed without winding the sub-track;
step 2.3: repeating the steps to traverse all the GPS track lines, and finally obtaining a group circle behavior track line set, wherein the extraction result of the group jogging circle track line set is shown in FIG. 4;
and step 3: as shown in fig. 5 and 6, geometric information and semantic information of the motion place are extracted from the colony circular trajectory line set extracted in step 2 by using methods such as Delaunay triangulation, inverse address coding, and the like.
Step 3.1: as shown in fig. 5(a) and 5(b), a constrained Delaunay triangulation network is constructed for the group circle trajectory line set extracted in step 2.3, and the whole long edge in the Delaunay triangulation network is deleted, so that the circle trajectory lines of each motion place are grouped into one type. The global long-edge deletion threshold GlobalValue is calculated as follows:
GlobalValue=Mean(DT)+α×Variation(DT)
mean (DT) in the above formula represents the average value of all side lengths of the triangular net DT; variance (DT) represents the standard deviation of all side lengths of the triangular net; alpha is a regulating parameter and is 1 by default. If the side length of any triangle side is greater than the GlobalValue, deleting the side, and the result of deleting the whole long side is shown in fig. 6 (b);
step 3.2: as shown in fig. 6(a), after the local long edge of the cluster obtained in step 3.1 is deleted again, a polygon is extracted by merging triangles for each cluster, and the polygon is simplified to obtain a motion path polygon. The local long edge deletion threshold LocalValue is calculated as follows:
in the above formulaRepresenting a cluster GiMidpoint pjThe average value of all side lengths in the second-order neighborhood of (1); LocalVarioration (p)j) Representing a cluster GiNeutralization point pjThe standard deviation of the side length of the direct connection side; beta is a regulating parameter, and the default is 2;
step 3.3: as shown in fig. 6(b), the inner ring is deleted from the path polygon extracted in step 3.2 to extract a circle-around place polygon, and then circle-around sport place semantic information is extracted by using Baidu reverse address coding, which mainly includes semantic types such as stadium, living cell, park, road, lake greenbelt, leisure greenbelt, greenbelt square, and the like.
It will be understood that modifications and variations can be made by persons skilled in the art in light of the above teachings and all such modifications and variations are intended to be included within the scope of the invention as defined in the appended claims.
Claims (7)
1. A method for extracting a sport place based on jogging GPS track data is characterized by comprising the following steps:
step 1, acquiring original jogging GPS trajectory data from a database, decomposing the original jogging GPS trajectory data into two-dimensional time series curve data, and extracting a key point set from the two-dimensional time series curve data;
step 2, constructing a triangular matrix at the sub-track distance according to the track point set corresponding to the key point set, and searching a winding sub-track meeting the winding track mode characteristics according to a line scanning search matrix; outputting a group winding sub-trajectory line set after traversing all GPS trajectory lines in the database;
the specific method of the step 2 comprises the following steps:
step 2.1, constructing a triangular matrix on the track distance by taking the SP point set obtained in the step 1.3 as a row and a column according to the time sequence, and recording the path distance (pathDist), the direction distance (dirDist) and the winding time length (T) of a sub-track formed by track points corresponding to the row and the column by each matrix unitdur) (ii) a For any trajectory T ═ pn,pn+1,...,pmThe calculation formulas of the track path distance, the direction distance and the duration are as follows:
dirDist(T)=Dist(pn,pm)
Tdur=T(pm)-T(pn)
wherein Dist (p)k,pk+1) Euclidean distance, T (p), representing two trace pointsk) Representing points of track pkThe time of recording;
step 2.2, searching the triangular matrix on the track distance according to rows, and if 3 eigenvalues in the matrix unit all meet the track characteristics of the winding behavior, the conditions are met:
{pathDist≥minPath&&dirDist≤maxDirct&&Tdur≥minTime}
the sub-track corresponding to the row and column is a winding sub-track, and the winding sub-tracks are sequentially searched according to rows by the matrix row where the terminal point of the sub-track is located; if the winding sub-track is searched, repeatedly executing the step 2.2 to perform depth-first search; outputting a winding action sub-track until the track line is processed completely without a winding action sub-track;
step 2.3, repeatedly executing the step 2.1 to the step 2.2 to traverse all GPS track lines, and finally obtaining a group circle behavior track set;
and 3, extracting geometric information and semantic information of the motion place from the group circle trajectory line set.
2. The jogging GPS trajectory data-based sport area extraction method according to claim 1, characterized in that the concrete method of step 1 is:
step 1.1, determining a parameter threshold, comprising: shortest path distance, maximum direction distance, minimum cycle time;
step 1.2, inputting original jogging GPS trajectory data, and decomposing original space-time trajectory coordinates (x, y, t) into two-dimensional time sequence curve data of (t, x) and (t, y);
step 1.3, taking a starting point, an end point, a maximum value and a minimum value point in the two-dimensional time sequence data as key points; merging the key point sets extracted from the (t, x) and (t, y) sequences to obtain a whole key point set KP ═ k1,k2,...,kn}; extracting original track points corresponding to the key points according to the time sequence to be used as a search point set SP ═ s of the winding behavior sub-track1,s2,...,sn}。
3. The jogging GPS track data-based sport area extraction method according to claim 2, characterized in that the parameter threshold determined in step 1.1 is specifically:
the shortest path distance minPath is 400m, the maximum direction distance maxDirct is 16m, and the minimum cycle time minTime is 5 min.
4. The jogging GPS track data-based sport area extraction method according to claim 2, wherein the method of selecting the extreme point in step 1.3 is specifically:
for a sequence of three points (x)i-1,xi,xi+1) If point xiSatisfy { xi<xi-1&&xi<xi+1Either { x } ori>xi-1&&xi>xi+1The pixel is a local extreme point; extreme point xiThe time period of the extreme value is kept, namely the time period of the point and the front and rear extreme points is greater than minTime/4, and the minTime represents the minimum cycle time; otherwise, the point is taken as a redundant key point and deleted.
5. The jogging GPS trajectory data-based sport field extraction method according to claim 1, characterized in that in step 3, Delaunay triangulation and inverse address coding are used to extract geometrical information and semantic information of sport fields from the colony circular trajectory set.
6. The jogging GPS track data-based sport area extraction method according to claim 1, characterized in that the specific method of step 3 is:
3.1, constructing a constrained Delaunay triangulation network for the group winding trajectory line set extracted in the step 2.3, and deleting the integral long edge in the Delaunay triangulation network, so that winding trajectory lines of each motion place are gathered into one type; the overall long-edge deletion threshold globalarue is calculated as follows:
GlobalValue=Mean(DT)+α×Variation(DT)
wherein mean (DT) represents the average value of all side lengths of the triangular net DT; variance (DT) represents the standard deviation of all side lengths of the triangular net; alpha is a regulating parameter and is 1 by default; for any triangle side, if the side length is larger than the GlobalValue, deleting the side;
step 3.2, after deleting the local long edge of the cluster obtained in the step 3.1 again, merging triangles for each cluster to extract polygons, and simplifying the polygons to obtain a motion path polygon; the local long edge deletion threshold LocalValue is calculated as follows:
wherein the content of the first and second substances,representing a cluster GiMidpoint pjThe average value of all side lengths in the second-order neighborhood of (1); LocalVarioration (p)j) Representing a cluster GiNeutralization point pjThe standard deviation of the side length of the direct connection side; beta is a regulating parameter, and the default is 2;
step 3.3, deleting an inner ring of the path polygon extracted in the step 3.2 to extract a circle place polygon, and then extracting circle motion place semantic information by utilizing Baidu reverse address coding, wherein the semantic types comprise: stadiums, living quarters, parks, roads, lake greens, leisure greens, greens squares.
7. The jogging GPS track data-based sports venue extraction method according to claim 1, wherein the database of step 1 records: the mass jogging GPS trajectory line data with individual marks is generated by the mobile positioning equipment and the sport APP.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810652492.0A CN108984643B (en) | 2018-06-22 | 2018-06-22 | Sport place extraction method based on jogging GPS track data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810652492.0A CN108984643B (en) | 2018-06-22 | 2018-06-22 | Sport place extraction method based on jogging GPS track data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108984643A CN108984643A (en) | 2018-12-11 |
CN108984643B true CN108984643B (en) | 2021-09-21 |
Family
ID=64538138
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810652492.0A Active CN108984643B (en) | 2018-06-22 | 2018-06-22 | Sport place extraction method based on jogging GPS track data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108984643B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110988880B (en) * | 2019-12-12 | 2021-09-07 | 南京莱斯电子设备有限公司 | Geographic information extraction and target tracking method based on SMR target track |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104573390A (en) * | 2015-01-27 | 2015-04-29 | 武汉大学 | Cognitive-rule-based time-space trajectory fusion method and road network topology generating method |
CN104700617A (en) * | 2015-04-02 | 2015-06-10 | 武汉大学 | High-precision lane information extracting method based on low-precision GPS track data |
EP2973254A1 (en) * | 2013-03-15 | 2016-01-20 | Conservis Corporation | Ticket-based harvest management system and method |
-
2018
- 2018-06-22 CN CN201810652492.0A patent/CN108984643B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2973254A1 (en) * | 2013-03-15 | 2016-01-20 | Conservis Corporation | Ticket-based harvest management system and method |
CN104573390A (en) * | 2015-01-27 | 2015-04-29 | 武汉大学 | Cognitive-rule-based time-space trajectory fusion method and road network topology generating method |
CN104700617A (en) * | 2015-04-02 | 2015-06-10 | 武汉大学 | High-precision lane information extracting method based on low-precision GPS track data |
Non-Patent Citations (1)
Title |
---|
"运用约束Delaunay三角网从众源轨迹线提取道路边界";杨伟等;《测绘学报》;20170228;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN108984643A (en) | 2018-12-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Guo et al. | Analysis of factors affecting urban park service area in Beijing: Perspectives from multi-source geographic data | |
Chen et al. | Predicting the effect of street environment on residents' mood states in large urban areas using machine learning and street view images | |
WO2020172954A1 (en) | Living circle identification method based on positioning data | |
Zhou et al. | A data-driven method for trip ends identification using large-scale smartphone-based GPS tracking data | |
Gong et al. | Crowd characterization for crowd management using social media data in city events | |
CN106156528A (en) | A kind of track data stops recognition methods and system | |
CN112819319A (en) | Method for measuring correlation between city vitality and spatial social characteristics and application | |
CN111666823B (en) | Pedestrian re-identification method based on individual walking motion space-time law collaborative identification | |
CN107729363A (en) | Based on GoogLeNet network model birds population identifying and analyzing methods | |
CN115512216A (en) | City functional area fine recognition method coupling block space-time characteristics and ensemble learning | |
Xia et al. | Decision tree-based contextual location prediction from mobile device logs | |
Zhou et al. | A pedestrian network construction system based on crowdsourced walking trajectories | |
Kisilevich et al. | A novel approach to mining travel sequences using collections of geotagged photos | |
CN112365708A (en) | Scenic spot traffic volume prediction model establishing and predicting method based on multi-graph convolution network | |
CN108984643B (en) | Sport place extraction method based on jogging GPS track data | |
CN111325153A (en) | Student behavior characteristic intelligent analysis method based on multidimensional data | |
Yue et al. | A visualized soundscape prediction model for design processes in urban parks | |
Luo et al. | FBVA: A flow-based visual analytics approach for citywide crowd mobility | |
Zhao et al. | Learning region similarities via graph-based deep metric learning | |
Zhang et al. | Urban Visual Intelligence: Studying Cities with AI and Street-level Imagery | |
CN115752459A (en) | Trajectory rectification method based on indoor position network model | |
Signorello et al. | Exploring geo-tagged photos to assess spatial patterns of visitors in protected areas: the case of park of Etna (Italy) | |
Tang et al. | RETRACTED ARTICLE: Distribution of earthquake activity in mountain area based on embedded system and physical fitness detection of basketball | |
Mao et al. | The greener the living environment, the better the health? Examining the effects of multiple green exposure metrics on physical activity and health among young students | |
Yang et al. | Cycle Periodic Behavior Detection and Sports Place Extraction Using Crowdsourced Running Trace Data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |